Skip to main content

Data migration services for Imago have certain image requirements for compatibility with AutoCrop ML services.  Imago AutoCrop capabilities can accurately determine the edges of whole core trays and individual core rows providing geologists with instant access to whole core tray images and core samples with the background omitted.

As part of the Imago data migration terms, all AutoCrop results will be manually verified, and extra ML training may be applied to suit the images if required. But if the migrated images do not follow the basic structural & layout standards shown below, Seequent cannot crop the images, even if retraining is done or not

Before committing to data migration services with AutoCrop, Seequent recommends checking a representative sample of images for compliance.


AutoCrop is only compatible with images of core trays where boxes are stacked contiguously and vertically aligned.

Angled alignment

Core trays must have the same alignment.  Core trays should run horizontally across the image, parallel to the top / bottom edge of the image.

Gaps between boxes

For images containing multiple core trays, boxes should have a gap no larger than any one core row diameter.

Incomplete core trays

AutoCrop is designed to ignore incomplete core trays.  This is an optimization for certain imaging stations where adjacent core trays sometimes enter the image frame.  Any migrated images containing partially obscured core trays will be ignored by AutoCrop.

Core rows

AutoCrop can identify up to 20 core tray rows per image.  AutoCrop’s accuracy is best when images hold less than 10 rows however.